Clinical Impact of an Early Response and of Early Initiation of Salvage Therapy in Patients with Systemic Light Chain (AL) Amyloidosis

A deep and rapid hematologic response is fundamental in order to improve the outcomes (overall survival (OS) and organ responses) of patients with AL amyloidosis. However, the optimal time to assess the efficacy of therapy and make decisions has not been established, with most data based on assessments performed after 3 months of therapy. However, given the toxicity of the circulating free light chains, a delay in response may have detrimental consequences, especially for high risk patients. In patients not optimally responding, early rather than late change in therapy may salvage some of them, but there is limited data. We aimed to evaluate the significance of an early response (i.e at 1 and at 3 months after start of therapy) and of an earlier introduction of salvage therapy, in patients not achieving at least a very good partial response (VGPR).

The analysis included consecutive previously untreated patients with AL amyloidosis, who received bortezomib-based primary therapy, in a single center (Department of Clinical Therapeutics, Athens, Greece). The analysis included 198 patients who had evaluable clonal disease (i.e dFLC>20 mg/L with abnormal FLC ratio, or serum monoclonal protein >0.5 gr/dl) and available assessments at 1 and at 3 months from the start of therapy, excluding patients who died early (<3 months) (44 patients excluded because of early death, 26 for non-evaluable disease and 14 for missing assessments).

The median age was 64 years (range 40-84), 57% were males. Heart was involved in 64%, per Mayo stage 23%, 52%, 18% and 7% were stage 1,2, 3A & 3B. Kidneys were involved in 72%, median eGFR was 71 ml/min/1.73 m2, median proteinuria was 5.3 gr/d; peripheral/autonomic nerve was involved in 18% and liver in 21%.

After 1 month of therapy, 38% had not achieved a response (NR), 35% had a VGPR and 27% a PR. Among patients who had <VGPR after 1 month of therapy (n=126), only 24% reached a VGPR by 3 months so that 25% had NR, 24% had a PR and 51% had a VGPR at 3 months.

We then assessed the impact of the depth of response at 1 and at 3 months. One-year OS of patients with VGPR, PR and NR at 1 month was 85%, 84% and 67% and at 3 years was 78%, 53% and 48% respectively (median OS: 9.5 vs 3.1 vs 2 years, p=0.001). According to the response at 3 months landmark, 1 year OS was 92% vs 69% vs 61% for VGPR, PR and NR respectively ( median OS: 7 vs 3.5 vs 1.2 years respectively, p<0.001). Patients that at 3 months improved their response to VGPR (from PR or NR) had an OS of 4.2 years vs 2.4 years for those that either remained in PR or improved their response from NR to PR while it was 1.6 years for those that had NR (vs 9.5 for those in VGPR at 1 month, overall p<0.001). The OS difference among those at VGPR at 1 month vs those that improved their response to VGPR at 3 months was also significant (p=0.031).

Patients that after 3 months of therapy had not achieved at least a VGPR were analyzed according to whether they remained on the same treatment or started salvage therapy immediately. Most received salvage therapy with lenalidomide/dexamethasone or with lenalidomide added to bortezomib. Of those with NR, 32% started a new therapy at this timepoint and 68% continued on the same therapy, while among those in PR, 80% continued the same therapy.

At 6 months, 37% (10% VGPR, 27% PR) of patients who had NR at 3 months and remained on the same therapy achieved a response vs 44% (20% VGPR, 24% PR) of those that switched to salvage therapy. Among those in PR at 3 months, only 12.5% of those continuing with the same therapy improved their response to VGPR. Patients with <VGPR at 3 months that continued their initial regimen had an OS of 41 vs 6 months for those that received salvage therapy at 3 months (p=0.009), probably because these were sicker (50% vs 12% were stage 3).

In multivariate analysis, at least VGPR after 1 month (p=0.001) or at 3 months (p=0.004) and Mayo stage 1 or 2 (vs stage 3) (p<0.001) were the stronger prognostic factors for survival at each respective landmark point.

In conclusion, in this cohort of bortezomib-treated patients with AL amyloidosis, a very early (at 1 month) and deep response is associated with better survival. Early introduction of salvage therapy may improve the depth of response and survival for some patients, but, with the available therapies, it may be too late for many patients even if salvage is given early. Based on our results, starting with the most effective therapy to achieve rapidly the deepest response possible, especially in patients at higher risk, is the best strategy.